[FieldTrip] Problems with filterpadding in ft_artifact_zvalue

[Brehm, J. (Julia)]

Dear list,

I am currently working on implementing an automatic artifact rejection using ft_artifact_zvalue.
It already works quite nicely for detecting jumps and EOG artifacts. However, when trying to detect high frequency noise I am often getting edge effects. I am already using filterpadding for the eog artifacts were it does a good job, but for the higher frequencies it suddenly almost only detects artifacts at the outermost trials (first and last) even though there are also obvious artifacts in different trials. The data is previously demeaned and filtered using a notch filter and a bandpass filter from 1-100 Hz. These are my settings for the artifact detection now:

cfg            = [];
cfg.trl        = temptrl; % all trials in this case

cfg.artfctdef.zvalue.channel     = channel{ch}; % single channel
cfg.artfctdef.zvalue.cutoff      = 3;
cfg.artfctdef.zvalue.fltpadding  = 0.1;

cfg.artfctdef.zvalue.bpfilter    = 'yes';
cfg.artfctdef.zvalue.bpfilttype = 'but';
cfg.artfctdef.zvalue.bpfreq      = [70 100];
cfg.artfctdef.zvalue.bpfiltord   = 2;
cfg.artfctdef.zvalue.hilbert     = 'yes';
cfg.artfctdef.zvalue.boxcar      = 0.2;

[cfg, artifact] = ft_artifact_zvalue(cfg, data);

I already tried changing the settings around but that didn't do much good. Any help would be appreciated and please tell me if I am doing anything terribly wrong!

Best,
Julia
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.science.ru.nl/pipermail/fieldtrip/attachments/20160425/ff230941/attachment-0001.html>